Title :
Reduced-Complexity Delayed-Decision Algorithm for Context-Based Image Processing Systems
Author :
Singh, Raghavendra ; Ortega, Antonio
Author_Institution :
IBM India Res. Lab., Delhi
Abstract :
It is well known that the performance of context-based image processing systems can be improved by allowing the processor (e.g., an encoder or a denoiser) a delay of several samples before making a processing decision. Often, however, for such systems, traditional delayed-decision algorithms can become computationally prohibitive due to the growth in the size of the space of possible solutions. In this paper, we propose a reduced-complexity, one-pass, delayed-decision algorithm that systematically reduces the size of the search space, while also preserving its structure. In particular, we apply the proposed algorithm to two examples of adaptive context-based image processing systems, an image coding system that employs a context-based entropy coder, and a spatially adaptive image-denoising system. For these two types of widely used systems, we show that the proposed delayed-decision search algorithm outperforms instantaneous-decision algorithms with only a small increase in complexity. We also show that the performance of the proposed algorithm is better than that of other, higher complexity, delayed-decision algorithms.
Keywords :
image coding; image denoising; adaptive image-denoising system; context-based entropy coder; context-based image processing systems; delayed-decision search algorithm; image coding system; processing decision; reduced-complexity delayed-decision algorithm; Adaptive systems; Added delay; Delay systems; Entropy coding; Image coding; Image denoising; Image processing; Image restoration; Laboratories; Quantization; Delayed-decision processing; image coding; image denoising; image processing; Algorithms; Artificial Intelligence; Data Compression; Decision Support Techniques; Image Enhancement; Image Interpretation, Computer-Assisted; Pattern Recognition, Automated; Reproducibility of Results; Sensitivity and Specificity;
Journal_Title :
Image Processing, IEEE Transactions on
DOI :
10.1109/TIP.2007.901246